Options
Adaptive Fuzzy CMAC Control for a Class of Nonlinear Systems
Date Issued
2006
Date
2006
Author(s)
Wu, Ter-Feng
DOI
en-US
Abstract
ABSTRACT
In this thesis, a modified multivariable adaptive fuzzy cerebellar model articulation controller (CMAC) control scheme is proposed to solve the tracking problem for a class of nonlinear systems. Firstly, a fuzzy CMAC (FCMAC) that merges fuzzy logic and CMAC algorithm such that the input space dimension and the complicated structure in CMAC can be simplified. The FCMAC module is used to approximate a nonlinear multivariable (multi-input multi-output (MIMO)) system involving uncertainty to create the desired ideal control inputs. Next, suitable control and adaptive laws with output feedback based on sliding surface concept are incorporated with FCMAC into a multi-input single-output (MISO) adaptive FCMAC (AFCMAC) control system, to tune all of the control gains on-line, thereby accommodating the uncertainty of nonlinear systems without prior off-line learning phase. Particularly, to reduce the approximation error, improve the tracking accuracy, and guarantee the closed-loop stability, the conventional switching robust compensation is adopted. Furthermore, to overcome the chattering problem associated with discontinuity derived from switching action, a smooth compensation is then proposed, completing the modified MISO AFCMAC control scheme.
Eventually, the theories and applications concerning the modified MISO AFCMAC control scheme is further to extend successfully to the modified MIMO AFCMAC control scheme as the main results of this work. By integrated Lyapunov stability analysis, it is guaranteed that all of the closed-loop signals are bounded, and the tracking errors converge exponentially to a residual set, whose size can be adjusted by changing the design parameters. On the whole, although the tracking precision is reduced slightly, the control signal’s quality can be improved greatly. Finally, simulation results for its applications to several examples are presented to demonstrate the validity and applicability of the methodologies proposed in this thesis.
In this thesis, a modified multivariable adaptive fuzzy cerebellar model articulation controller (CMAC) control scheme is proposed to solve the tracking problem for a class of nonlinear systems. Firstly, a fuzzy CMAC (FCMAC) that merges fuzzy logic and CMAC algorithm such that the input space dimension and the complicated structure in CMAC can be simplified. The FCMAC module is used to approximate a nonlinear multivariable (multi-input multi-output (MIMO)) system involving uncertainty to create the desired ideal control inputs. Next, suitable control and adaptive laws with output feedback based on sliding surface concept are incorporated with FCMAC into a multi-input single-output (MISO) adaptive FCMAC (AFCMAC) control system, to tune all of the control gains on-line, thereby accommodating the uncertainty of nonlinear systems without prior off-line learning phase. Particularly, to reduce the approximation error, improve the tracking accuracy, and guarantee the closed-loop stability, the conventional switching robust compensation is adopted. Furthermore, to overcome the chattering problem associated with discontinuity derived from switching action, a smooth compensation is then proposed, completing the modified MISO AFCMAC control scheme.
Eventually, the theories and applications concerning the modified MISO AFCMAC control scheme is further to extend successfully to the modified MIMO AFCMAC control scheme as the main results of this work. By integrated Lyapunov stability analysis, it is guaranteed that all of the closed-loop signals are bounded, and the tracking errors converge exponentially to a residual set, whose size can be adjusted by changing the design parameters. On the whole, although the tracking precision is reduced slightly, the control signal’s quality can be improved greatly. Finally, simulation results for its applications to several examples are presented to demonstrate the validity and applicability of the methodologies proposed in this thesis.
Subjects
適應控制
模糊控制
小腦模式控制
智慧型控制
非線性系統
Adaptive Control
Fuzzy Control
Cerebellar Model Articulation Controller (CMAC)
Intelligent Control
Nonlinear Systems
Type
thesis
File(s)
No Thumbnail Available
Name
ntu-95-D86921015-1.pdf
Size
23.31 KB
Format
Adobe PDF
Checksum
(MD5):801c85520956d8aa859e31fc7ada0a52